Algorithm and Examples of an Agent-Based Evacuation Model

Abstract

This research establishes a “detect-decide-action” agent-based evacuation model based on the social force model, introducing an active steering force into the basis of the dynamic equation with the combination of the behavioral decision model and the probability model. In the AEM, the detection algorithm is used to identify pedestrians or obstacles within the detection radius to provide the next walking direction and apply the active steering force. The obstacle avoidance algorithm is the core of the “action” link. This research focuses on the establishment of the following and bypassing algorithm when moving in the same direction, and the algorithm of a detour when moving in the opposite direction, applying C++ programming language to achieve the basic evacuation behavior simulation of avoiding pedestrians and obstacles in the actual scene. The results show that compared with the grid model and the general social force model, the agent model (AEM) solves the problem of the distortion of evacuation behavior to some extent, and the pedestrian is more flexible in the choice of evacuation path.